Covariate effects on the apparent clearance of tacrolimus in paediatric liver transplant patients undergoing conversion therapy

Objective: To analyse the influence of covariates on the apparent clearance (CL) of tacrolimus in paediatric liver transplant recipients being convert ed from cyclosporin to tacrolimus. Design: Retrospective modelling study. Patients and participants: 18 children, 13 girls and 5 boys, aged 4 months to 16 years (median 9.1 years) who required conversion to tacrolimus becaus e of acute or chronic rejection or cyclosporin toxicity. Methods: 287 whole-blood tacrolimus concentrations from therapeutic drug mo nitoring were used to build a nonlinear mixed-effects population model (NON -MEM program) for the apparent clearance of tacrolimus. Variables considere d were age, total bodyweight (TBW), body surface area (BSA), time after ini tiation of treatment (T), gender, haematocrit (Hct), albumin (Alb), asparta te aminotransferase (AST), alanine aminotransferase (ALT), gamma -glutamyl transpeptidase (gamma GT), alkaline phosphatase (ALP), bilirubin (BIL), cre atinine clearance (CLCR) and dosage of concomitant corticosteroids (EST). Results: TBW, T, BIL and ALT were the covariates that displayed a significa nt influence on CL according to the final regression model: CL (L/h) = 10.4 (TBW/70)3/4 . e(-0.00032) T. e(-0.057) (BIL.) (1 - 0.079 ALT). With this mo del, the estimates of the coefficients of variation were 24.3% and 29.5% fo r interpatient variability in CL and residual variability, respectively. Conclusions: The proposed model for tacrolimus CL can be applied for a prio ri dosage calculations, although the results should be used with caution be cause of the unexplained variability in the CL. We therefore recommended cl ose monitoring of tacrolimus whole blood concentrations, especially within the first months of treatment. The best use of the model would be its appli cation in dosage adjustment based on therapeutic drug monitoring and the Ba yesian approach.